Skip to content

Productivity

Productivity answers one question: is what your agents do worth it? It shows how many tasks met their goal, how long the agents actually worked, the work they handled in your systems, and — for owners — what each goal met cost. Everything on it is measured; nothing is a “time saved” guess.

It is a page of the admin area at aptiveai.io/admin/productivity, the Productivity row in the admin menu. In AptiveAI Agents, owners and admins also find a Productivity card on the Work page.

RoleSees Productivity
Owner✓ — your own enterprise, with AI usage in dollars
Admin✓ when they administer every workspace of the enterprise — without any money: no spend tile, no Is it worth it? chart, no AI usage columns. An admin of only some workspaces does not see enterprise-wide figures
Member— Productivity is visible to your enterprise’s owners and admins.

The page uses the same time range as Activity, with Refresh in the header. Tasks are counted the same way: tasks started in the period, deleted tasks left out.

TileWhat it shows
Goals metGoals met ÷ all tasks started in the period. The hint gives N of M tasks, what task status alone says — the share that simply completed — and how many were not graded
Agent work timeHow long the agents actually worked on these tasks, with the average per task. Time spent waiting for a person, a question or an approval adds nothing
Spent per goal metOwners only. The AI usage of all the period’s tasks ÷ goals met, with that AI usage in the hint
Work handledRecords read — distinct records your connectors returned to the agents — with connector actions, approved actions and pages visited

Goals met and task status alone says sit side by side on purpose. A task can complete and still not do what was asked; the gap between the two is the work that looked done and was not. What a goal met is, and when a task is graded, is explained on Activity.

A task with no verdict never counts as met, so many ungraded tasks pull the rate down rather than inflate it. A line under the tiles says how many tasks were not graded, and why that happens: grading is off, the workspace is regulated, or the grade has not arrived yet.

  • Task outcomes — each period’s tasks: Goal met, Partly met, Blocked, Not met, Stopped and Not graded.
  • Why tasks didn’t meet the goal — what stood in the way of tasks that did not meet their goal (missing information, a missing file, a sign-in, the live view, not allowed, a system error, out of scope, something else), and how ungraded tasks ended (stopped by a person, an approval declined or timed out, AI usage ran out). No task is counted twice.
  • Is it worth it? — Owners only. AI usage per period; point at a bar for what each goal met cost, with a line of the cost per goal met over the periods that had one.

One row per agent: Tasks, Goals met, Success (goals met ÷ tasks), Work time, work time Per task, Records read, Connector actions and Approved actions. Owners also see AI usage and AI usage Per goal met.

For one agent at a time, in AptiveAI Agents, open its dashboard: its tasks, goals met, time worked and files, counted over the tasks you can see.